Accurate single-molecule spot detection for image-based spatial transcriptomics with weakly supervised deep learning

• Published in: bioRxiv
• Main Authors: Laubscher, Emily, Wang, Xuefei, Razin, Nitzan, Dougherty, Tom, Xu, Rosalind, Lincoln Ombelets, Pao, Edward, Moffitt, Jeffrey, Yue, Yisong, Van Valen, David Ashley
• Format: Paper
• Language: English
• Published: Cold Spring Harbor Cold Spring Harbor Laboratory Press 05.02.2024
• Subjects: Cell culture; Deep learning; Gene expression; Image processing; Spatial discrimination learning; Transcriptomes; Transcriptomics
• DOI: 10.1101/2023.09.03.556122

Image-based spatial transcriptomics methods enable transcriptome-scale gene expression measurements with spatial information but require complex, manually-tuned analysis pipelines. We present Polaris, an analysis pipeline for image-based spatial transcriptomics that combines deep learning models for cell segmentation and spot detection with a probabilistic gene decoder to quantify single-cell gene expression accurately. Polaris is available through the DeepCell software library (https://github.com/vanvalenlab/deepcell-spots) and https://www.deepcell.org.Competing Interest StatementDVV is a co-founder and Chief Scientist of Barrier Biosciences and holds equity in the company. DVV, EL, and NR filed a patent for weakly supervised deep learning for spot detection. JRM is co-founder and scientific advisor to Vizgen and holds equity in the company. JRM is an inventor on patents related to MERFISH filed on his behalf by Harvard University and Boston Children's Hospital. All other authors declare no competing interests.Footnotes* The manuscript has been updated to include the airlocalize spot detection method as an additional classical spot detection method.* https://deepcell.readthedocs.io/en/master/data-gallery/index.html